Moonshots with Peter Diamandis

EP #197: The Latest in AI — Job Loss, Elon & Sam Altman Chip Race & the "AI Bubble"

Guests: Brian Elliott (Blitzy), Emad Mostaque (Intelligent Internet), Dave Blundin (Link Exponential Ventures)
Date: September 26, 2025

Overview

In this episode, Peter Diamandis and co-hosts explore the relentless acceleration of artificial intelligence, its wide-reaching impact on jobs and education, and the trillion-dollar race for AI compute infrastructure. Special guests Brian Elliott, Emad Mostaque, and Dave Blundin offer insider commentary on the "AI wars" (Google, Microsoft, OpenAI, Meta, xAI), the transforming landscape of higher education, the ongoing data center arms race, the coming disruption in healthcare and capital markets, and the underlying question: is this an AI bubble — or a true economic inflection point?

Key Discussion Points & Insights

1. Is AI a Bubble?

• Consensus: Absolutely not. The panel draws contrasts to the dotcom era, arguing that AI produces real, rapidly growing economic value.
• Dave Blundin: "Nothing's going to change the world more than what's going on right now. It is definitely not a bubble. It's not even vaguely like a bubble." (00:06)
• Imad Mostaque: "AI is useful. People pay for it because it has economic value." (00:00)
• Data Evidence: Nvidia's meteoric rise matches up with earnings growth, not just speculation. CapEx transforms immediately into increased output via AI applications. (74:58–79:21)

2. The Higher Education Crisis

• Context: Value perception of college has plummeted. In 2010, 75% of Americans said college was "very important". In 2025, only 35% agree; meanwhile, "not too important" responses have soared.
• Peter Diamandis: "The cost of tuition goes through the roof. The perceived value of the education has been plummeting ... you're saddled with debt and you don't make it back because you're not getting the jobs. It's crazy." (03:08)
• Brian Elliott: Credentials matter more than content/learning; dropping out of an elite school can sometimes get you funded faster than graduating.
• Imad Mostaque: The "AI drop" is accelerating, with early-stage graduates struggling to find employment — and AI amplifying this trend. (04:41–05:26)
• Systemic forces: Top universities are shielded by massive endowments; second- and third-tier schools face existential risk.
• Future vision: "We haven't seen the first AI university yet, which I think is going to be really interesting..." (09:24)

3. AI Agents and Job Displacement

• Impact: Rapid automation threatens knowledge and skilled labor. The only "future-proof" path may be entrepreneurship or domain-specific innovation using AI.
• Peter Diamandis: "The only career of the future that really matters... is being an entrepreneur. It's not marching up the career path." (05:26)
• Dave Blundin: Immediate productivity gains can lead to staff reduction, especially in fast-growing companies (e.g., Duolingo, Salesforce). But many CEOs and public figures are now dampening their messaging to avoid panic. (86:45)
• Imad Mostaque: "Humans will have negative value in cognitive labor in a few years ... if you're in a job where your role is to beat other people, as in private sector competitive jobs, you're not going to beat an AI." (83:00)
• Real-world stats: GPT4 outperforms physicians at diagnosis 92% vs. 74% for humans; adding a human lowers performance. (84:15)

4. The Compute Arms Race: Chips, Data Centers, and Power

• Stakes: Google, OpenAI, xAI, Meta, and others are racing to secure rare compute, build immense data centers, and secure energy—described as unprecedented capital war.
• Highlights:
  • xAI (Elon Musk) pledges to hit 10 gigawatts of compute, then 100GW, then 1 terawatt: "It's basically like as much as a state." (19:26)
  • OpenAI, via Nvidia partnership, throwing $100 billion at compute, planning multi-square-mile "superbrain" campuses. (20:00–21:26)
  • Scarcity: Annual chip production is far less than anticipated demand; even dominant players must maneuver, hedge, and optimize to maintain access.
  • Compute is power: "We're converting electrons into intelligence and into crypto." (22:24)
• Key quote: Greg Brockman: "By 2030, we will be in a world of material abundance... except for compute, which will be absolutely scarce." (27:13–27:31)

5. AI Model Wars: Who Wins?

• Gemini (Google) overtaking ChatGPT in US iOS installs, thanks to Google's distribution muscle. (12:29–12:51)
• Alibaba’s Qwen model rising due to relentless iteration and massive data.
• Grok5 (xAI) leads AGI benchmarks, e.g., ARC AGI v2, New York Times Connections—though there's suspicion of "benchmaxing" by training specifically to optimize for each test. (14:38–18:38)
• Brian Elliott: Cost per task matters as much as model intelligence: scalable, cheap compute will determine the winner. (16:05)

6. Vertical Integration: Hardware, Models, Applications

• Google: Vertically integrated with TPUs, no Nvidia tax, five times more power efficient for certain AI tasks.
• Meta (Zuckerberg): Committed $600 billion to US data centers by 2028: "Better to lose billions than be late to superintelligence." (35:22)
• Microsoft & OpenAI: In a high-stake dance/trade for compute and ownership control, with OpenAI aiming to restructure into a for-profit and spawn a record-breaking nonprofit endowment. (32:12–34:49)

7. Economic Transformation & Tokenized Securities

• Tokenized securities, 24/7 trading: NASDAQ and Robinhood leading the charge to on-chain finance—stocks, and even privates like OpenAI and SpaceX, soon to be tokenized and tradeable. (88:05–93:16)
• Dave Blundin: Blockchains and AI can replace outdated public market regulations, reporting, and liquidity demands, enabling smaller, faster IPO-like structures.

8. Healthcare and Drug Discovery Revolution

• Wearables: Apple Watch receives FDA clearance for hypertension detection; CGM, Oura, "insideables" feed actionable personal health data to AI.
• AI-accelerated drug discovery:
  • First AI-designed drugs hit clinical trials; process cut from years to months.
  • 150+ new small molecule drugs discovered by AI in 2025; phase 1 success rates 80–90%, far above historical averages. (96:46–97:56)
  • Emad Mostaque predicts the FDA will soon demand in silico (AI-predicted) trial results.

Notable Quotes & Memorable Moments

• On AI Bubble vs. Reality:
  "AI is useful. People pay for it because it has economic value." – Imad Mostaque (00:00)

• On Higher Education’s Obsolescence:
  "We can build a nuclear reactor on campus faster than we will ever change this curriculum." – Dave Blundin (03:03)

• On Compute Scarcity as Bottleneck:
  "Abundance everywhere except compute scarcity... fundamentally, it is a physical infrastructure problem, not just a software problem." – Greg Brockman (27:13–27:31)

• On Opportunity:
  "There will never be a better time. There never has been a better time. Maybe tomorrow." – Dave Blundin (98:57)

• On Human Value in Work:
  "Humans will have negative value in cognitive labor in a few years... If you’re in a job where your role is to beat other people... you’re not going to beat an AI." – Imad Mostaque (83:00)

• On The Infrastructure Armistice:
  "We’re converting electrons into intelligence and into, into crypto." – Peter Diamandis (22:24)

Timestamps for Key Segments

• Intro & Main Theme

  • Is AI a Bubble? (00:00–00:25)

• Education Crisis & AI

  • College Value Plummets (01:21–10:20)
  • The AI University Paradigm (09:24–10:20)

• AI Wars and Compute Race

  • Gemini, ChatGPT, Alibaba Qwen, Grok5* (12:29–19:00)
  • xAI, OpenAI, Meta Data Center Race (19:00–23:24)
  • Physical limits of chip production, scarcity (21:26–25:50)

• AI’s Impact on the Labor Market

  • Unemployment and automation (05:26–08:23, 81:39–86:45)
  • Three-day workweek? The end of human cognitive labor (81:39–85:22)
  • Domain Experts and New Entrepreneurship (47:14–53:58)

• Hardware & Integration

  • Google, TPUs, Vertically Integrated AI (44:39–46:11)

• Structural Changes: Economy & Finance

  • Tokenized Securities, Blockchain IPOs (88:05–93:36)
  • Abundance, Immediate Payback vs. Bubble (74:22–79:21)

• Health & Drug Discovery

  • Wearable’s FDA Cleared for Hypertension (93:36–96:46)
  • AI-accelerated Drug Discovery (96:46–98:18)

Final Takeaways

• Relentless Acceleration: AI progress, economic impact, and infrastructure buildout are all compounding more quickly than any previous tech transformation — and show every sign of continuing exponential growth.
• Labor Market Disruption: Entrepreneurs and domain experts who leverage AI are best positioned, while traditional credentialing and middle-skilled knowledge work are increasingly at risk.
• AI Will Touch Everything: From university reform to tokenized capital markets, from autonomous “wearables” for health to robots in homes and factories, the platforms being built now are laying the groundwork for a vastly different society.
• Now is the Window: As Dave Blundin and Brian Elliott both stress, there may never be a better time for bold action, experimentation, learning, and investing in this technological wave.

For Further Exploration

• Slides & Resources: diamandis.com/wtf
• Newsletter (Metatrends): dashmandis.com/metatrends
• Intelligent Internet Book: thelasteconomy.com

*For a full, detailed transcript and more resources, visit the podcast website or subscribe to Peter Diamandis' newsletter.